
# Import required libraries and pandas to read the raw data csv file to a dataframe

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

import seaborn as sns
%matplotlib inline


# Read the consolidated csv file into pandas dataframe 
RAW_DATASET = pd.read_csv('data/ConsolidatedDataV2.csv')


# Print the number of rows and columns in the raw dataset
RAW_DATASET.shape

(2856, 30)


# Print the first 6 rows of the raw dataset
RAW_DATASET.head()


# Print the last 6 rows of the raw dataset
RAW_DATASET.tail()


# Generate descriptive statistics on raw dataset
RAW_DATASET.describe()


# Print the data types of the columns in raw dataset
RAW_DATASET.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 2856 entries, 0 to 2855
Data columns (total 30 columns):
 #   Column               Non-Null Count  Dtype  
---  ------               --------------  -----  
 0   Country              2856 non-null   object 
 1   Year                 2856 non-null   int64  
 2   Life expectancy      2567 non-null   float64
 3   BMI                  2856 non-null   float64
 4   ChildMalnutrition    2822 non-null   float64
 5   Cholera              781 non-null    float64
 6   Alcohol              2800 non-null   float64
 7   HIV                  2040 non-null   float64
 8   BCG                  2344 non-null   float64
 9   Adult Mortality      2856 non-null   int64  
 10  ChildMortality       2856 non-null   int64  
 11  Population           2528 non-null   float64
 12  Eggs Consumption     2339 non-null   float64
 13  Bovine Meat          2339 non-null   float64
 14  Mutton & Goat meat   2339 non-null   float64
 15  Other Meat           2339 non-null   float64
 16  Pig Meat             2273 non-null   float64
 17  Poultry Meat         2339 non-null   float64
 18  Milk Consumption     2339 non-null   float64
 19  Fish and Seafood     2339 non-null   float64
 20  Medical Expenditure  2458 non-null   float64
 21  Retirement Age       782 non-null    float64
 22  Diphtheria           2837 non-null   float64
 23  Suicides             2856 non-null   float64
 24  NCD                  2856 non-null   float64
 25  Env Pollution        2856 non-null   float64
 26  HepatitisB           2288 non-null   float64
 27  Measles              1629 non-null   float64
 28  Polio                2820 non-null   float64
 29  Tuberculosis         2799 non-null   float64
dtypes: float64(26), int64(3), object(1)
memory usage: 669.5+ KB


# Creatine new dataframe called "dataset" to process and store the raw data 
dataset = RAW_DATASET


# The dataset consists of “Year” column which is time series data. 

# TODO : Change data type of "Year" column to datetime
dataset['Year'] = pd.to_datetime(dataset['Year'] , format='%Y', errors='ignore')
dataset.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 2856 entries, 0 to 2855
Data columns (total 30 columns):
 #   Column               Non-Null Count  Dtype         
---  ------               --------------  -----         
 0   Country              2856 non-null   object        
 1   Year                 2856 non-null   datetime64[ns]
 2   Life expectancy      2567 non-null   float64       
 3   BMI                  2856 non-null   float64       
 4   ChildMalnutrition    2822 non-null   float64       
 5   Cholera              781 non-null    float64       
 6   Alcohol              2800 non-null   float64       
 7   HIV                  2040 non-null   float64       
 8   BCG                  2344 non-null   float64       
 9   Adult Mortality      2856 non-null   int64         
 10  ChildMortality       2856 non-null   int64         
 11  Population           2528 non-null   float64       
 12  Eggs Consumption     2339 non-null   float64       
 13  Bovine Meat          2339 non-null   float64       
 14  Mutton & Goat meat   2339 non-null   float64       
 15  Other Meat           2339 non-null   float64       
 16  Pig Meat             2273 non-null   float64       
 17  Poultry Meat         2339 non-null   float64       
 18  Milk Consumption     2339 non-null   float64       
 19  Fish and Seafood     2339 non-null   float64       
 20  Medical Expenditure  2458 non-null   float64       
 21  Retirement Age       782 non-null    float64       
 22  Diphtheria           2837 non-null   float64       
 23  Suicides             2856 non-null   float64       
 24  NCD                  2856 non-null   float64       
 25  Env Pollution        2856 non-null   float64       
 26  HepatitisB           2288 non-null   float64       
 27  Measles              1629 non-null   float64       
 28  Polio                2820 non-null   float64       
 29  Tuberculosis         2799 non-null   float64       
dtypes: datetime64[ns](1), float64(26), int64(2), object(1)
memory usage: 669.5+ KB


# Print the first 6 rows of the dataset
dataset.head()


# Total columns = 30
# Total rows = 2856

# TODO : check the count of missing values by columns in descending order
dataset.isnull().sum().sort_values(ascending=False)

Cholera                2075
Retirement Age         2074
Measles                1227
HIV                     816
Pig Meat                583
HepatitisB              568
Eggs Consumption        517
Milk Consumption        517
Mutton & Goat meat      517
Bovine Meat             517
Other Meat              517
Fish and Seafood        517
Poultry Meat            517
BCG                     512
Medical Expenditure     398
Population              328
Life expectancy         289
Tuberculosis             57
Alcohol                  56
Polio                    36
ChildMalnutrition        34
Diphtheria               19
Year                      0
ChildMortality            0
Adult Mortality           0
Suicides                  0
NCD                       0
Env Pollution             0
BMI                       0
Country                   0
dtype: int64


# TODO : Drop the columns with more than 50% of missing data 

# Cholera                2075
# Retirement Age         2074
# Measles                1227

# Drop the cholera column more than 50% missing data 
dataset = dataset.drop(['Cholera'],axis=1)

# Drop the retitrement column 
dataset = dataset.drop(['Retirement Age'],axis=1)

# Drop the Measles column
dataset = dataset.drop(['Measles'],axis=1)


# Life Expectancy is the main focus of our research
# Imputing introduces bias which is not desirable . Hence, we delete this null rows 
# Life expectancy         289

# Filter the rows with null values "Life expectancy" column of dataset
dataset=dataset[dataset['Life expectancy'].notna()]


# Convert datatype of life expectancy to int after removing null values
# To apply classification algorithms

dataset['Life expectancy'] = dataset['Life expectancy'].astype(np.int64)


# TODO : check the count of missing values by columns in descending order after removing some columns and rows
dataset.isnull().sum().sort_values(ascending=False)

HIV                    731
HepatitisB             522
BCG                    476
Pig Meat               294
Mutton & Goat meat     228
Other Meat             228
Fish and Seafood       228
Milk Consumption       228
Poultry Meat           228
Eggs Consumption       228
Bovine Meat            228
Medical Expenditure    187
Population             124
Alcohol                 56
Tuberculosis            38
ChildMalnutrition       34
Polio                   34
Diphtheria              17
Year                     0
ChildMortality           0
Adult Mortality          0
BMI                      0
Suicides                 0
NCD                      0
Env Pollution            0
Life expectancy          0
Country                  0
dtype: int64


# TODO : Drop only the rows with missing values which are less than 10% 

# Pig Meat               294
# Mutton & Goat meat     228
# Other Meat             228
# Fish and Seafood       228
# Milk Consumption       228
# Poultry Meat           228
# Eggs Consumption       228
# Bovine Meat            228
# Medical Expenditure    187
# Population             124
# Alcohol                 56
# Tuberculosis            38
# ChildMalnutrition       34
# Polio                   34
# Diphtheria              17

dataset=dataset[dataset['Pig Meat'].notna()]
dataset=dataset[dataset['Mutton & Goat meat'].notna()]
dataset=dataset[dataset['Other Meat'].notna()]

dataset=dataset[dataset['Fish and Seafood'].notna()]
dataset=dataset[dataset['Milk Consumption'].notna()]
dataset=dataset[dataset['Poultry Meat'].notna()]

dataset=dataset[dataset['Eggs Consumption'].notna()]
dataset=dataset[dataset['Bovine Meat'].notna()]
dataset=dataset[dataset['Medical Expenditure'].notna()]

dataset=dataset[dataset['Population'].notna()]
dataset=dataset[dataset['Alcohol'].notna()]
dataset=dataset[dataset['Tuberculosis'].notna()]

dataset=dataset[dataset['ChildMalnutrition'].notna()]
dataset=dataset[dataset['Polio'].notna()]
dataset=dataset[dataset['Diphtheria'].notna()]


# TODO : Replace the missing values with mean value
# missing data is about 25% and deletion causes too much data loss from the dataset ( rows decrease drastically)

# HIV                     731
# HepatitisB              522
# BCG                     476

dataset['HIV']=dataset['HIV'].fillna(value=dataset['HIV'].mean())
dataset['BCG']=dataset['BCG'].fillna(value=dataset['BCG'].mean())
dataset['HepatitisB']=dataset['HepatitisB'].fillna(value=dataset['HepatitisB'].mean())


# Ignore and supress the warnings thrown by dataprep library 
import warnings
warnings.filterwarnings('ignore')


# Import the dataprep library
import dataprep

from dataprep.eda import create_report

# Generate the visual report called "pre_processed_report" using create_report method
pre_processed_report = create_report(dataset, title='Pre-Processed Dataset')
pre_processed_report

NumExpr defaulting to 8 threads.


# Using label encoder for the categorical columns with text values
from sklearn.preprocessing import LabelEncoder
le = LabelEncoder()

# TODO : Encode labels "Country" and "Year"
dataset['Country'] = le.fit_transform(dataset['Country'])
dataset['Year'] = le.fit_transform(dataset['Year'])
dataset.head()


# Plot the correleation between all the features in the dataset 
# Identify the strongly related related variables with Life Expectancy

#Using Pearson Correlation
plt.figure(figsize=(20,15))
correlation_matrix = dataset.corr()
sns.heatmap(correlation_matrix, annot=True, vmin=-1, vmax=1, center= 0, cmap= 'coolwarm')
plt.show()


# Correlation with output variable Life expectancy
cor_target = correlation_matrix["Life expectancy"]

# View highly correlated features
relevant_features = cor_target[cor_target>=-1]
relevant_features.sort_values(ascending=False)

Life expectancy        1.000000
Eggs Consumption       0.708943
Milk Consumption       0.659305
BMI                    0.654637
Polio                  0.649462
Diphtheria             0.641525
Pig Meat               0.591232
Poultry Meat           0.590431
Alcohol                0.450114
Bovine Meat            0.439087
Medical Expenditure    0.364535
Fish and Seafood       0.330970
HepatitisB             0.244216
BCG                    0.235935
Year                   0.183378
Mutton & Goat meat     0.056031
Population             0.010906
Country               -0.067061
Other Meat            -0.090164
Tuberculosis          -0.106261
ChildMortality        -0.250402
HIV                   -0.326044
Suicides              -0.429654
NCD                   -0.706556
Env Pollution         -0.708975
ChildMalnutrition     -0.882725
Adult Mortality       -0.948039
Name: Life expectancy, dtype: float64


# Drop features with no strong correlation with the target variable from the dataset 

# HepatitisB             0.244216
# BCG                    0.235935
# Year                   0.183378
# Mutton & Goat meat     0.056031
# Population             0.010906
# Country               -0.067061
# Other Meat            -0.090164
# Tuberculosis          -0.106261

dataset=dataset.drop('HepatitisB',axis=1)
dataset=dataset.drop('BCG',axis=1)
dataset=dataset.drop('Year',axis=1)
dataset=dataset.drop('Mutton & Goat meat',axis=1)
dataset=dataset.drop('Population',axis=1)
dataset=dataset.drop('Country',axis=1)
dataset=dataset.drop('Other Meat',axis=1)
dataset=dataset.drop('Tuberculosis',axis=1)


# Print first 6 rows in dataset 
dataset.head()


# TODO : Feature Scaling

#transform the data to be on same scale using sklearn's StandardScaler()
from sklearn.preprocessing import StandardScaler

scale = StandardScaler()


X = dataset.drop('Life expectancy',axis=1)
y = dataset['Life expectancy'].astype('int')

# Print the features "X" and target variable "y"
print( X , y)

       BMI  ChildMalnutrition  Alcohol      HIV  Adult Mortality  \
0     26.2               22.9     5.03    100.0              103   
1     26.3               23.1     4.43    100.0              103   
2     26.4               23.6     4.28    100.0              100   
3     25.4               33.9     0.54    200.0               98   
4     25.5               33.9     0.55    200.0               96   
...    ...                ...      ...      ...              ...   
2851  23.7               48.3     3.91  40000.0              440   
2852  23.7               46.2     3.93  33000.0              407   
2853  23.7               44.3     4.11  28000.0              383   
2854  23.8               42.8     4.22  25000.0              358   
2855  23.8               41.7     3.84  24000.0              346   

      ChildMortality  Eggs Consumption  Bovine Meat  Pig Meat  Poultry Meat  \
0                868              7.72        21.24     11.03         13.41   
1                868             12.69        22.40     11.04         12.76   
2                868             12.45        22.50     10.88         13.23   
3              60319              7.93         5.43      0.00          6.86   
4              60319              8.65         5.35      0.00          6.64   
...              ...               ...          ...       ...           ...   
2851           52218              1.78         7.66      2.81          6.69   
2852           52218              1.78         7.53      2.88          6.72   
2853           52218              1.75         7.37      2.65          4.97   
2854           52218              1.93         7.26      1.84          4.49   
2855           52218              1.84         7.25      1.74          4.67   

      Milk Consumption  Fish and Seafood  Medical Expenditure  Diphtheria  \
0               301.27              5.86             4.795327        99.0   
1               299.85              4.97             5.055262        99.0   
2               303.72              4.87             5.385599        99.0   
3               151.06              4.40             6.547214        95.0   
4               125.37              4.16             6.978492        95.0   
...                ...               ...                  ...         ...   
2851             34.21              2.99             8.081738        93.0   
2852             31.07              2.91             6.918353        95.0   
2853             31.90              2.82             7.110148        95.0   
2854             30.54              3.39             8.133524        91.0   
2855             27.38              3.82             7.452066        87.0   

      Suicides    NCD  Env Pollution  Polio  
0          7.6  490.2            0.5   99.0  
1          4.8  507.1            0.4   99.0  
2          4.8  486.4            0.4   99.0  
3          2.8  464.4            0.8   95.0  
4          2.7  460.7            0.7   95.0  
...        ...    ...            ...    ...  
2851      34.3  842.2            4.6   93.0  
2852      33.1  826.4            4.5   95.0  
2853      31.4  810.2            4.5   95.0  
2854      30.8  804.3            4.4   92.0  
2855      30.7  800.1            4.3   88.0  

[2074 rows x 18 columns] 0       76
1       77
2       77
3       75
4       76
        ..
2851    52
2852    55
2853    56
2854    58
2855    59
Name: Life expectancy, Length: 2074, dtype: int32


X = scale.fit_transform(X)
X

array([[ 0.69850001, -0.60006687, -0.04445178, ..., -0.54018371,
        -0.62734346,  0.83874704],
       [ 0.75026479, -0.59081548, -0.19637236, ..., -0.45311988,
        -0.69201418,  0.83874704],
       [ 0.80202956, -0.567687  , -0.2343525 , ..., -0.5597602 ,
        -0.69201418,  0.83874704],
       ...,
       [-0.59561939,  0.38983192, -0.27739667, ...,  1.10836225,
         1.95948557,  0.56552109],
       [-0.54385461,  0.32044649, -0.24954456, ...,  1.07796719,
         1.89481484,  0.36060162],
       [-0.54385461,  0.26956384, -0.34576093, ...,  1.05633002,
         1.83014412,  0.08737566]])


#splitting the data into my train and test set 
from sklearn.model_selection import train_test_split

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3,random_state=0)


#from sklearn import utils
#print(utils.multiclass.type_of_target(y_train))
#print(utils.multiclass.type_of_target(y_train.astype('int')))

y_train = y_train.astype('int')
y_test = y_test.astype('int')


np.unique(y_train, return_counts=True)

(array([39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55,
        56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72,
        73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84]),
 array([  1,   1,   1,   5,   6,   7,   7,   9,  12,  13,  12,  24,  24,
         19,  19,  17,  21,  20,  33,  28,  27,  24,  27,  26,  26,  18,
         24,  37,  38,  33,  46,  57,  65,  62,  71, 107,  74,  47,  36,
         66,  76,  70,  72,  37,   5,   1], dtype=int64))


from imblearn.over_sampling import RandomOverSampler

ros = RandomOverSampler(random_state=0)
X_train,y_train = ros.fit_resample(X, y)

np.unique(y_train, return_counts=True)

(array([39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55,
        56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72,
        73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84]),
 array([137, 137, 137, 137, 137, 137, 137, 137, 137, 137, 137, 137, 137,
        137, 137, 137, 137, 137, 137, 137, 137, 137, 137, 137, 137, 137,
        137, 137, 137, 137, 137, 137, 137, 137, 137, 137, 137, 137, 137,
        137, 137, 137, 137, 137, 137, 137], dtype=int64))


from sklearn.metrics import r2_score, mean_absolute_error
from sklearn.metrics import mean_squared_error
from sklearn.metrics import confusion_matrix
from sklearn.metrics import f1_score
from math import sqrt


from sklearn.linear_model import LogisticRegression

#defining default values to store high accuracy and values of C at higher ccuracy
HIGH_ACCURACY =0
MAX_C = 1

# Looping for values of c from 1 till 20 to check the behaviour and accuracy of the model.
for c in range(1,20):
    
    #defining linear model with default parameters and C=c
    model = LogisticRegression(penalty='l2', dual=False, tol=0.0001, C=c, 
                           fit_intercept=True, intercept_scaling=1, class_weight=None, 
                           random_state=None, solver='lbfgs', max_iter=100, multi_class='auto', 
                           verbose=0, warm_start=False, n_jobs=None, l1_ratio=None)
    
    #Fitting the model for the test and train sets
    model = model.fit(X_train,y_train)

    #Predicting the test values with the built model 
    pred_y = model.predict(X_test)
    
    #printing the accuracy on the console
    print("C = {} ,  Accuracy = {}".format(c, model.score(X_test, y_test)))
    
    #if accuracy is high then store it in the variables
    if(HIGH_ACCURACY < model.score(X_test, y_test)):
        HIGH_ACCURACY = model.score(X_test, y_test)
        MAX_C = c
        
    
#Printing the the highest accuracy achieved and the respective C value     
print("*************************************")
print("Maximum Accuracy is achieved at C = {} ,  Accuracy = {}".format(MAX_C, HIGH_ACCURACY))

C = 1 ,  Accuracy = 0.3467094703049759
C = 2 ,  Accuracy = 0.36436597110754415
C = 3 ,  Accuracy = 0.3611556982343499
C = 4 ,  Accuracy = 0.3611556982343499
C = 5 ,  Accuracy = 0.36276083467094705
C = 6 ,  Accuracy = 0.36597110754414125
C = 7 ,  Accuracy = 0.36757624398073835
C = 8 ,  Accuracy = 0.36757624398073835
C = 9 ,  Accuracy = 0.36918138041733545
C = 10 ,  Accuracy = 0.3723916532905297
C = 11 ,  Accuracy = 0.3739967897271268
C = 12 ,  Accuracy = 0.36918138041733545
C = 13 ,  Accuracy = 0.36597110754414125
C = 14 ,  Accuracy = 0.38202247191011235
C = 15 ,  Accuracy = 0.3723916532905297
C = 16 ,  Accuracy = 0.36918138041733545
C = 17 ,  Accuracy = 0.3739967897271268
C = 18 ,  Accuracy = 0.36757624398073835
C = 19 ,  Accuracy = 0.36757624398073835
*************************************
Maximum Accuracy is achieved at C = 14 ,  Accuracy = 0.38202247191011235


# DEFINING MODEL WITH PARAMETER C=14 (HIGH ACCURACY)
model = LogisticRegression(C=14)

# TRAINING THE DEFINED MODEL
model = model.fit(X_train,y_train)

# PREDICTING THE TARGET WITH THE TRAINED MODEL
y_pred = model.predict(X_test)

# CALCULATING MEAN ABSOLUTE ERROR
MAE = mean_absolute_error(y_test, y_pred)

# CALCULATING THE MEAN SQUARED ERROR
MSE = mean_squared_error(y_test, y_pred)

# CALCULATING ROOT MEAN SQUARED ERROR
RMSE = sqrt(MSE)

# CALCULATING R2_SCORE 
R2_SCORE=r2_score(y_test, y_pred)

# CALCULATING F1_SCORE 
F1_SCORE = f1_score(y_test, y_pred, average='macro')

#PRINTING THE CALCULATED METRICS
print('LOGISTIC REGRESSION , C=14')
print('------------------------------')
# PRININTING THE ACCURACY OF THE MODEL WITH DEFINED C VALUE
print('Accuracy : {}'.format(model.score(X_test, y_test)))
print('MAE : {}'.format(round(MAE, 2)))
print('MSE : {}'.format(round(MSE, 2)))
print('RMSE  : %f' % RMSE)
print('R2_SCORE  : %f' % R2_SCORE)
print('F1_SCORE  : %f' % F1_SCORE)

LOGISTIC REGRESSION , C=14
------------------------------
Accuracy : 0.38202247191011235
MAE : 1.16
MSE : 3.83
RMSE  : 1.958232
R2_SCORE  : 0.960430
F1_SCORE  : 0.340216


# GAUSSIAN NAIVE BAYES

#IMPORTING GAUSSIAN NAIVE BAYES PACKAGE
from sklearn.naive_bayes import GaussianNB

# DEFINING MODEL
gnb = GaussianNB()

# TRAINING THE DEFINED MODEL AND PREDICTING THE TARGET WITH TRAINED MODEL
y_pred = gnb.fit(X_train, y_train).predict(X_test)

# CALCULATING MEAN ABSOLUTE ERROR
MAE = mean_absolute_error(y_test, y_pred)

# CALCULATING THE MEAN SQUARED ERROR
MSE = mean_squared_error(y_test, y_pred)

# CALCULATING ROOT MEAN SQUARED ERROR
RMSE = sqrt(MSE)

# CALCULATING R2_SCORE 
R2_SCORE=r2_score(y_test, y_pred)

# CALCULATING F1_SCORE 
F1_SCORE = f1_score(y_test, y_pred, average='macro')

print('GAUSSIAN NAIVE BAYES ')
print('------------------------------')
print('Accuracy : {}'.format(gnb.score(X_test, y_test)))
print('MAE : {}'.format(round(MAE, 2)))
print('MSE : {}'.format(round(MSE, 2)))
print('RMSE  : %f' % RMSE)
print('R2_SCORE  : %f' % R2_SCORE)
print('F1_SCORE  : %f' % F1_SCORE)

GAUSSIAN NAIVE BAYES 
------------------------------
Accuracy : 0.30818619582664525
MAE : 1.63
MSE : 5.96
RMSE  : 2.440956
R2_SCORE  : 0.938516
F1_SCORE  : 0.249275


# BernoulliNB Naive Bayes
from sklearn.naive_bayes import BernoulliNB

# DEFINING MODEL
bnb = BernoulliNB()

# TRAINING THE DEFINED MODEL AND PREDICTING THE TARGET WITH TRAINED MODEL
model = bnb.fit(X_train, y_train).predict(X_test)

# CALCULATING MEAN ABSOLUTE ERROR
MAE = mean_absolute_error(y_test, y_pred)

# CALCULATING THE MEAN SQUARED ERROR
MSE = mean_squared_error(y_test, y_pred)

# CALCULATING ROOT MEAN SQUARED ERROR
RMSE = sqrt(MSE)

# CALCULATING R2_SCORE 
R2_SCORE=r2_score(y_test, y_pred)

# CALCULATING F1_SCORE 
F1_SCORE = f1_score(y_test, y_pred, average='macro')

print('BERNOULLI NAIVE BAYES ')
print('------------------------------')
print('Accuracy : {}'.format(bnb.score(X_test, y_test)))
print('MAE : {}'.format(round(MAE, 2)))
print('MSE : {}'.format(round(MSE, 2)))
print('RMSE  : %f' % RMSE)
print('R2_SCORE  : %f' % R2_SCORE)
print('F1_SCORE  : %f' % F1_SCORE)

BERNOULLI NAIVE BAYES 
------------------------------
Accuracy : 0.22150882825040127
MAE : 1.63
MSE : 5.96
RMSE  : 2.440956
R2_SCORE  : 0.938516
F1_SCORE  : 0.249275


from sklearn.neighbors import KNeighborsClassifier

#defining default values to store high accuracy and values of n at higher ccuracy
HIGH_ACCURACY =0
MAX_N = 1

# Looping for values of n from 5 till 20 to check the behaviour and accuracy of the model.
for n in range(5,20):
    
    #defining model with default parameters and n_neighbors = n
    knn = KNeighborsClassifier(n_neighbors = n)
    
    #Fitting the model for the test and train sets
    knn.fit(X_train,y_train)
    
     #Predicting the test values with the built model
    y_pred = knn.predict(X_test)
    
    #printing the accuracy on the console
    print('n = {} , Accuracy = {}'.format(n, knn.score(X_test, y_test)))
    
     #if accuracy is high then store it in the variables
    if(HIGH_ACCURACY < knn.score(X_test, y_test)):
        HIGH_ACCURACY = knn.score(X_test, y_test)
        MAX_N= n
        
    
# print('KNeighborsClassifier')
#Printing the the highest accuracy achieved and the respective C value     
print("*************************************")
print("Maximum Accuracy is achieved at n = {} ,  Accuracy = {}".format(MAX_N, HIGH_ACCURACY))

n = 5 , Accuracy = 0.8507223113964687
n = 6 , Accuracy = 0.797752808988764
n = 7 , Accuracy = 0.7736757624398074
n = 8 , Accuracy = 0.7223113964686998
n = 9 , Accuracy = 0.6934189406099518
n = 10 , Accuracy = 0.666131621187801
n = 11 , Accuracy = 0.6324237560192616
n = 12 , Accuracy = 0.5858747993579454
n = 13 , Accuracy = 0.565008025682183
n = 14 , Accuracy = 0.5280898876404494
n = 15 , Accuracy = 0.5136436597110754
n = 16 , Accuracy = 0.4767255216693419
n = 17 , Accuracy = 0.47191011235955055
n = 18 , Accuracy = 0.47351524879614765
n = 19 , Accuracy = 0.4510433386837881
*************************************
Maximum Accuracy is achieved at n = 5 ,  Accuracy = 0.8507223113964687


print('KNeighborsClassifier')
print('------------------------------')

# initialising the classifier for n=5
knn = KNeighborsClassifier(n_neighbors = 5)

# applying the model for the test values
knn.fit(X_train,y_train)

# predicting the out put values for test inputs 
y_pred = knn.predict(X_test)

print('Accuracy : {}'.format(knn.score(X_test, y_test)))

MAE = mean_absolute_error(y_test, y_pred)
print('MAE : {}'.format(round(MAE, 2)))

MSE = mean_squared_error(y_test, y_pred)
print('MSE : {}'.format(round(MSE, 2)))

RMSE = sqrt(MSE)
print('RMSE  : %f' % RMSE)

R2_SCORE=r2_score(y_test, y_pred)
print('R2_SCORE  : %f' % R2_SCORE)

F1_SCORE = f1_score(y_test, y_pred, average='macro')
print('F1_SCORE  : %f' % F1_SCORE)

KNeighborsClassifier
------------------------------
Accuracy : 0.8507223113964687
MAE : 0.17
MSE : 0.21
RMSE  : 0.458555
R2_SCORE  : 0.997830
F1_SCORE  : 0.892837


#SVC, NuSVC and LinearSVC perform both binary and multi-class classification on a dataset.


#Importing the necessary packages and libaries
from sklearn import svm

print('SUPPORT VECTOR CLASSIFICATION ')

# LIST OF KERNALS
kernal = ["linear","rbf",'poly','sigmoid']

#LOOPING THROUGH THE LIST OF KERNALS
for k in kernal:
   
    print('------------------------------')
    
    # LOOPING THROUGH THE PARAMETER g TO TUNE OUR MODEL ACCORDINGLY 
    for g in ["auto" ,"scale"]:
        HIGH_ACC = 0
        Max_C = 1
        
        # LOOPING THROUGH VALUE OF C FROM 1 TO 15
        # higher value of c gives l2 penality --> overfitting
        for c in range(1,15):
            
            # DEFINING THE MODEL
            model = svm.SVC(C=c, kernel=k, gamma= g, decision_function_shape='ovo')
            
            # TRAINING THE CREATED MODEL
            model = model.fit(X_train,y_train)
            
            # PREDICTING THE TATRGET WITH TEST VALUES
            pred_y = model.predict(X_test)
            
            # STORING THE HIGH ACCURACY AND CORRESPONDING C AND g VALUES
            if HIGH_ACC < model.score(X_test,y_test):
                HIGH_ACC = model.score(X_test,y_test)
                Max_C = c

        # PRINTING THE HIGH ACCURACY FOR ALL THE KERNAL WITH DIFF PARAMETER COMBINATIONS
        print("Kernal = {} , C = {} , gamma = {} - MaxAccuracy = {}".format(k,Max_C,g,HIGH_ACC))

SUPPORT VECTOR CLASSIFICATION 
------------------------------
Kernal = linear , C = 10 , gamma = auto - MaxAccuracy = 0.6773675762439807
Kernal = linear , C = 10 , gamma = scale - MaxAccuracy = 0.6773675762439807
------------------------------
Kernal = rbf , C = 14 , gamma = auto - MaxAccuracy = 0.8154093097913323
Kernal = rbf , C = 14 , gamma = scale - MaxAccuracy = 0.7479935794542536
------------------------------
Kernal = poly , C = 14 , gamma = auto - MaxAccuracy = 0.7897271268057785
Kernal = poly , C = 14 , gamma = scale - MaxAccuracy = 0.6821829855537721
------------------------------
Kernal = sigmoid , C = 4 , gamma = auto - MaxAccuracy = 0.22632423756019263
Kernal = sigmoid , C = 5 , gamma = scale - MaxAccuracy = 0.28892455858747995


print('Linear SVC')
print('------------------------------')

# initialising the classifier
linear = svm.SVC(kernel='linear', C=10, decision_function_shape='ovo')

# applying the model for the test values
linear.fit(X_train,y_train)

# predicting the out put values for test inputs 
y_pred = linear.predict(X_test)

print('Accuracy : {}'.format(linear.score(X_test, y_test)))

MAE = mean_absolute_error(y_test, y_pred)
print('MAE : {}'.format(round(MAE, 2)))

MSE = mean_squared_error(y_test, y_pred)
print('MSE : {}'.format(round(MSE, 2)))

RMSE = sqrt(MSE)
print('RMSE  : %f' % RMSE)

R2_SCORE=r2_score(y_test, y_pred)
print('R2_SCORE  : %f' % R2_SCORE)

F1_SCORE = f1_score(y_test, y_pred, average='macro')
print('F1_SCORE  : %f' % F1_SCORE)

Linear SVC
------------------------------
Accuracy : 0.6773675762439807
MAE : 0.4
MSE : 0.58
RMSE  : 0.763325
R2_SCORE  : 0.993987
F1_SCORE  : 0.703947


print('Radial Basis Function - SVC')
print('------------------------------')

# initialising the classifier
rbf = svm.SVC(kernel='rbf', gamma="auto", C=14, decision_function_shape='ovo')

# applying the model for the test values
rbf.fit(X_train,y_train)

# predicting the out put values for test inputs 
y_pred = rbf.predict(X_test)

print('Accuracy : {}'.format(rbf.score(X_test, y_test)))

MAE = mean_absolute_error(y_test, y_pred)
print('MAE : {}'.format(round(MAE, 2)))

MSE = mean_squared_error(y_test, y_pred)
print('MSE : {}'.format(round(MSE, 2)))

RMSE = sqrt(MSE)
print('RMSE  : %f' % RMSE)

R2_SCORE=r2_score(y_test, y_pred)
print('R2_SCORE  : %f' % R2_SCORE)

F1_SCORE = f1_score(y_test, y_pred, average='macro')
print('F1_SCORE  : %f' % F1_SCORE)

Radial Basis Function - SVC
------------------------------
Accuracy : 0.8154093097913323
MAE : 0.21
MSE : 0.29
RMSE  : 0.541978
R2_SCORE  : 0.996969
F1_SCORE  : 0.812503


# polynomial kernel function 

poly = svm.SVC(kernel='poly', degree=3, C=14, gamma="auto", decision_function_shape='ovo').fit(X_train, y_train)

print('Polynomial Kernal Function - SVC')
print('------------------------------')

# predicting the out put values for test inputs 
y_pred = rbf.predict(X_test)

print('Accuracy : {}'.format(poly.score(X_test, y_test)))

MAE = mean_absolute_error(y_test, y_pred)
print('MAE : {}'.format(round(MAE, 2)))

MSE = mean_squared_error(y_test, y_pred)
print('MSE : {}'.format(round(MSE, 2)))

RMSE = sqrt(MSE)
print('RMSE  : %f' % RMSE)

R2_SCORE=r2_score(y_test, y_pred)
print('R2_SCORE  : %f' % R2_SCORE)

F1_SCORE = f1_score(y_test, y_pred, average='macro')
print('F1_SCORE  : %f' % F1_SCORE)

Polynomial Kernal Function - SVC
------------------------------
Accuracy : 0.7897271268057785
MAE : 0.21
MSE : 0.29
RMSE  : 0.541978
R2_SCORE  : 0.996969
F1_SCORE  : 0.812503


#Sigmoid
sig = svm.SVC(kernel='sigmoid', C=5, gamma="scale", decision_function_shape='ovo').fit(X_train, y_train)


print('Sigmoid Function - SVC')
print('------------------------------')

# predicting the out put values for test inputs 
y_pred = rbf.predict(X_test)

print('Accuracy : {}'.format(sig.score(X_test, y_test)))

MAE = mean_absolute_error(y_test, y_pred)
print('MAE : {}'.format(round(MAE, 2)))

MSE = mean_squared_error(y_test, y_pred)
print('MSE : {}'.format(round(MSE, 2)))

RMSE = sqrt(MSE)
print('RMSE  : %f' % RMSE)

R2_SCORE=r2_score(y_test, y_pred)
print('R2_SCORE  : %f' % R2_SCORE)

F1_SCORE = f1_score(y_test, y_pred, average='macro')
print('F1_SCORE  : %f' % F1_SCORE)

Sigmoid Function - SVC
------------------------------
Accuracy : 0.28892455858747995
MAE : 0.21
MSE : 0.29
RMSE  : 0.541978
R2_SCORE  : 0.996969
F1_SCORE  : 0.812503


# MODEL                      Accuracy       MAE     MSE      RMSE      R2_Score    F1_SCORE
# Gaussian Naive Bayes         31%          1.63    5.96   2.440956    0.938516    0.249275
# Bernoulli Naive Bayes        22%          2.95    23.42  4.839478    0.758323    0.153571
# Logistic Regression          38%          1.16    3.83   1.958232    0.96043     0.340216
# KNN Classifier               85%          0.17    0.21   0.458555    0.99783     0.892837
# Linear SVC                   68%          0.4     0.58   0.763325    0.993987    0.703947
# Radial Basis Function SVC    82%          0.21    0.29   0.541978    0.996969    0.812503
# Polynomial SVC               79%          0.21    0.29   0.541978    0.996969    0.812503
# Sigmoid SVC                  29%          0.21    0.29   0.541978    0.996969    0.812503



# prefer Knn because of higher F1score , lower rmse , higher r2 score . 
# Though accuracy is not a measure to choose right model. KNN stands to be the best model for the current dataset.


# To ignore warnings thrown by dataprep library 
import warnings
warnings.filterwarnings('ignore')

# importing the dataprep library
import dataprep
from dataprep.eda import create_report

# generating the visual using create_report method
final_report = create_report(dataset, title='Final Dataset')
final_report


#creating a new dataset with life expectancy greater than 73
countries_above73 = RAW_DATASET[RAW_DATASET['Life expectancy']>73]


# Listing the countries with high consumption of egg
countries_above73[countries_above73['Eggs Consumption']>18]['Country'].unique()

array(['China', 'Mexico', 'Netherlands', 'Japan', 'Denmark', 'Kuwait',
       'Paraguay'], dtype=object)


# Listing the countries with high consumption of beef
countries_above73[countries_above73['Bovine Meat']>35]['Country'].unique()

array(['Argentina', 'Australia', 'Brazil', 'New Zealand', 'Uruguay'],
      dtype=object)


# Listing the countries with high consumption of Pork
countries_above73[countries_above73['Pig Meat']>50]['Country'].unique()

array(['Poland', 'Montenegro', 'Spain', 'Germany', 'Austria',
       'Luxembourg', 'Netherlands'], dtype=object)


# Listing the countries with high consumption of Poultry Meat
countries_above73[countries_above73['Poultry Meat']>50]['Country'].unique()

array(['Kuwait', 'Barbados', 'Israel', 'Jamaica', 'United Arab Emirates',
       'Trinidad and Tobago'], dtype=object)


# Listing the countries with high consumption of Milk
countries_above73[countries_above73['Milk Consumption']>300]['Country'].unique()

array(['Albania', 'Netherlands', 'Montenegro', 'Sweden', 'Estonia',
       'Lithuania', 'Switzerland', 'Denmark', 'Finland', 'Greece',
       'Ireland', 'Luxembourg'], dtype=object)


# Listing the countries with high consumption of seafood
countries_above73[countries_above73['Fish and Seafood']>50]['Country'].unique()

array(['Malaysia', 'Maldives', 'Iceland', 'Japan', 'Norway', 'Portugal'],
      dtype=object)


# Grouping protein consumption columns 
# Protein_intake = Eggs Consumption + Bovine Meat + Pig Meat + Poultry Meat + Poultry Meat + Fish and Seafood

dataset["Protein_Intake"] = dataset["Eggs Consumption"] + dataset["Bovine Meat"] + dataset["Pig Meat"] + dataset["Poultry Meat"] + dataset["Poultry Meat"] + dataset["Fish and Seafood"]

dataset.head()


plt.rcParams["figure.figsize"] = (15,5)
plt.scatter(dataset["Life expectancy"], dataset["Protein_Intake"], alpha=0.5)
plt.xlabel('Life expectancy')
plt.ylabel('Protein_Intake')
plt.show()

	Year	Life expectancy	BMI	ChildMalnutrition	Cholera	Alcohol	HIV	BCG	Adult Mortality	ChildMortality	...	Medical Expenditure	Retirement Age	Diphtheria	Suicides	NCD	Env Pollution	HepatitisB	Measles	Polio	Tuberculosis
count	2856.000000	2567.000000	2856.000000	2822.000000	781.000000	2800.000000	2040.000000	2344.000000	2856.000000	2.856000e+03	...	2458.000000	782.000000	2837.000000	2856.000000	2856.000000	2856.000000	2288.000000	1629.000000	2820.000000	2799.000000
mean	2008.000000	68.795037	24.562185	36.663466	83.691421	4.814093	8918.289216	89.593003	194.384804	6.725880e+04	...	6.079373	66.125614	86.272823	11.992017	605.322024	1.498704	85.366696	84.666053	86.427660	8995.936395
std	4.899837	9.863356	3.390756	21.199494	301.232338	3.939412	24522.502650	13.897708	116.830456	2.234386e+05	...	2.435629	9.043176	15.585910	9.989647	186.165083	1.588911	17.253674	18.385550	15.209758	43737.466927
min	2000.000000	39.441000	0.000000	5.800000	0.000000	0.000000	100.000000	16.000000	49.000000	1.800000e+01	...	1.025159	57.035300	19.000000	0.100000	240.400000	0.000000	2.000000	2.000000	8.000000	0.000000
25%	2004.000000	61.433500	23.000000	19.500000	0.000000	1.240000	200.000000	87.000000	104.000000	1.062500e+03	...	4.152656	61.876000	82.000000	5.800000	456.575000	0.300000	81.000000	81.000000	82.000000	67.050000
50%	2008.000000	71.710000	25.400000	30.400000	4.000000	4.015000	975.000000	95.000000	164.000000	8.016500e+03	...	5.835874	64.249950	93.000000	10.000000	604.500000	0.800000	92.000000	92.000000	93.000000	646.000000
75%	2012.000000	76.129000	26.400000	52.475000	38.000000	7.752500	5600.000000	98.000000	260.000000	5.138925e+04	...	7.939192	67.866425	97.000000	15.000000	725.600000	2.400000	96.000000	97.000000	97.000000	3107.800000
max	2016.000000	84.090000	29.600000	88.400000	3990.000000	17.870000	290000.000000	99.000000	697.000000	2.025425e+06	...	20.413412	126.126400	99.000000	116.200000	1317.700000	9.400000	99.000000	99.000000	99.000000	615058.000000

Number of Variables	27
Number of Rows	2074
Missing Cells	0
Missing Cells (%)	0.0%
Duplicate Rows	0
Duplicate Rows (%)	0.0%
Total Size in Memory	568.7 KB
Average Row Size in Memory	280.8 B

Categorical	1
DateTime	1
Numerical	25

Distinct Count	125
Unique (%)	6.0%
Missing	0
Missing (%)	0.0%
Memory Size	147.5 KB

Mean	7.8038
Standard Deviation	3.1386
Median	7
Minimum	4
Maximum	24

	Country	Year	Life expectancy	BMI	ChildMalnutrition	Cholera	Alcohol	HIV	BCG	Adult Mortality	...	Medical Expenditure	Retirement Age	Diphtheria	Suicides	NCD	Env Pollution	HepatitisB	Measles	Polio	Tuberculosis
0	Albania	2011	76.914	26.2	22.9	NaN	5.03	100.0	97.0	103	...	4.795327	NaN	99.0	7.6	490.2	0.5	99.0	99.0	99.0	9.29
1	Albania	2012	77.252	26.3	23.1	NaN	4.43	100.0	96.0	103	...	5.055262	NaN	99.0	4.8	507.1	0.4	99.0	99.0	99.0	9.29
2	Albania	2013	77.554	26.4	23.6	NaN	4.28	100.0	99.0	100	...	5.385599	NaN	99.0	4.8	486.4	0.4	99.0	99.0	99.0	8.29
3	Algeria	2014	75.878	25.4	33.9	NaN	0.54	200.0	99.0	98	...	6.547214	NaN	95.0	2.8	464.4	0.8	95.0	99.0	95.0	3107.90
4	Algeria	2015	76.090	25.5	33.9	NaN	0.55	200.0	99.0	96	...	6.978492	NaN	95.0	2.7	460.7	0.7	95.0	99.0	95.0	3208.10

	Country	Year	Life expectancy	BMI	ChildMalnutrition	Cholera	Alcohol	HIV	BCG	Adult Mortality	...	Medical Expenditure	Retirement Age	Diphtheria	Suicides	NCD	Env Pollution	HepatitisB	Measles	Polio	Tuberculosis
2851	Zimbabwe	2011	52.896	23.7	48.3	2.0	3.91	40000.0	98.0	440	...	8.081738	NaN	93.0	34.3	842.2	4.6	94.0	NaN	93.0	1512.0
2852	Zimbabwe	2012	55.032	23.7	46.2	1.0	3.93	33000.0	98.0	407	...	6.918353	NaN	95.0	33.1	826.4	4.5	97.0	NaN	95.0	1612.0
2853	Zimbabwe	2013	56.897	23.7	44.3	NaN	4.11	28000.0	95.0	383	...	7.110148	NaN	95.0	31.4	810.2	4.5	95.0	NaN	95.0	1209.3
2854	Zimbabwe	2014	58.410	23.8	42.8	NaN	4.22	25000.0	99.0	358	...	8.133524	NaN	91.0	30.8	804.3	4.4	91.0	NaN	92.0	1410.0
2855	Zimbabwe	2015	59.534	23.8	41.7	0.0	3.84	24000.0	90.0	346	...	7.452066	NaN	87.0	30.7	800.1	4.3	87.0	NaN	88.0	1107.9

	Country	Year	Life expectancy	BMI	ChildMalnutrition	Cholera	Alcohol	HIV	BCG	Adult Mortality	...	Medical Expenditure	Retirement Age	Diphtheria	Suicides	NCD	Env Pollution	HepatitisB	Measles	Polio	Tuberculosis
0	Albania	2011-01-01	76.914	26.2	22.9	NaN	5.03	100.0	97.0	103	...	4.795327	NaN	99.0	7.6	490.2	0.5	99.0	99.0	99.0	9.29
1	Albania	2012-01-01	77.252	26.3	23.1	NaN	4.43	100.0	96.0	103	...	5.055262	NaN	99.0	4.8	507.1	0.4	99.0	99.0	99.0	9.29
2	Albania	2013-01-01	77.554	26.4	23.6	NaN	4.28	100.0	99.0	100	...	5.385599	NaN	99.0	4.8	486.4	0.4	99.0	99.0	99.0	8.29
3	Algeria	2014-01-01	75.878	25.4	33.9	NaN	0.54	200.0	99.0	98	...	6.547214	NaN	95.0	2.8	464.4	0.8	95.0	99.0	95.0	3107.90
4	Algeria	2015-01-01	76.090	25.5	33.9	NaN	0.55	200.0	99.0	96	...	6.978492	NaN	95.0	2.7	460.7	0.7	95.0	99.0	95.0	3208.10

1st row	Albania
2nd row	Albania
3rd row	Albania
4th row	Algeria
5th row	Algeria

Count	15929
Lowercase Letter	13633
Space Separator	239
Uppercase Letter	2296
Dash Punctuation	17
Decimal Number	0

Distinct Count	17
Unique (%)	0.8%
Missing	0
Missing (%)	0.0%
Memory Size	32.4 KB
Minimum	2000-01-01 00:00:00
Maximum	2016-01-01 00:00:00

Distinct Count	46
Unique (%)	2.2%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	68.7083
Minimum	39
Maximum	84
Zeros	0
Zeros (%)	0.0%
Negatives	0
Negatives (%)	0.0%

Minimum	39
5-th Percentile	49.65
Q1	62
Median	71
Q3	76
95-th Percentile	81
Maximum	84
Range	45
IQR	14

Distinct Count	94
Unique (%)	4.5%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	24.8506
Minimum	19.8
Maximum	29.1
Zeros	0
Zeros (%)	0.0%
Negatives	0
Negatives (%)	0.0%

Distinct Count	652
Unique (%)	31.4%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	35.8725
Minimum	9.3
Maximum	88.4
Zeros	0
Zeros (%)	0.0%
Negatives	0
Negatives (%)	0.0%

Mean	68.7083
Standard Deviation	10.0026
Variance	100.0523
Sum	142501
Skewness	-0.7547
Kurtosis	-0.3761
Coefficient of Variation	0.1456

Mean	24.8506
Standard Deviation	1.9323
Variance	3.7337
Sum	51540.2
Skewness	-0.5303
Kurtosis	-0.7565
Coefficient of Variation	0.07776

Mean	35.8725
Standard Deviation	21.6236
Variance	467.5794
Sum	74399.5
Skewness	0.7696
Kurtosis	-0.6317
Coefficient of Variation	0.6028

Mean	5.2056
Standard Deviation	3.9504
Variance	15.6055
Sum	10796.33
Skewness	0.5139
Kurtosis	-0.6963
Coefficient of Variation	0.7589

Minimum	9.3
5-th Percentile	11.7
Q1	18.8
Median	29.25
Q3	50.5
95-th Percentile	77.7
Maximum	88.4
Range	79.1
IQR	31.7

Distinct Count	1006
Unique (%)	48.5%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	5.2056
Minimum	0
Maximum	17.87
Zeros	25
Zeros (%)	1.2%
Negatives	0
Negatives (%)	0.0%

Minimum	0
5-th Percentile	0.19
Q1	1.68
Median	4.485
Q3	8.1
95-th Percentile	12.1
Maximum	17.87
Range	17.87
IQR	6.42

Distinct Count	219
Unique (%)	10.6%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	9416.6428
Minimum	100
Maximum	290000
Zeros	0
Zeros (%)	0.0%
Negatives	0
Negatives (%)	0.0%

Distinct Count	76
Unique (%)	3.7%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	89.9038
Minimum	16
Maximum	99
Zeros	0
Zeros (%)	0.0%
Negatives	0
Negatives (%)	0.0%

Distinct Count	445
Unique (%)	21.5%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	193.2985
Minimum	49
Maximum	683
Zeros	0
Zeros (%)	0.0%
Negatives	0
Negatives (%)	0.0%

Minimum	49
5-th Percentile	62.65
Q1	101
Median	161
Q3	255
95-th Percentile	445.35
Maximum	683
Range	634
IQR	154

Mean	75008.4851
Standard Deviation	254538.4035
Variance	6.479e+10
Sum	1.5557e+08
Skewness	6.1166
Kurtosis	39.4687
Coefficient of Variation	3.3935

Distinct Count	2073
Unique (%)	100.0%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	4.3879e+07
Minimum	247315
Maximum	1.3787e+09
Zeros	0
Zeros (%)	0.0%
Negatives	0
Negatives (%)	0.0%

Distinct Count	1105
Unique (%)	53.3%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	6.6446
Minimum	0.01
Maximum	22.35
Zeros	0
Zeros (%)	0.0%
Negatives	0
Negatives (%)	0.0%

LIFE EXPECTANCY PROJECT¶

Author : Mohana Kamanooru¶

Date : June 2021¶

1. Data Capture / Loading¶

2. Exploratory Data Analysis¶

3. Data Pre-processing¶

4. Data Visualisation¶

Visualising the data using dataprep library¶

Overview

Dataset Statistics

Variable Types

Variables

Country

Length

Sample

Letter

Year

Life expectancy

Quantile Statistics

Descriptive Statistics

BMI

Quantile Statistics

Descriptive Statistics

ChildMalnutrition

Quantile Statistics

Descriptive Statistics

Alcohol

Quantile Statistics

Descriptive Statistics

HIV

Quantile Statistics

Descriptive Statistics

BCG

Quantile Statistics

Descriptive Statistics

Adult Mortality

Quantile Statistics

Descriptive Statistics

ChildMortality

Quantile Statistics

Descriptive Statistics

Population

Quantile Statistics

Descriptive Statistics

Eggs Consumption

Quantile Statistics

Descriptive Statistics

Bovine Meat

Quantile Statistics

Descriptive Statistics

Mutton & Goat meat

Quantile Statistics

Descriptive Statistics

Other Meat

Quantile Statistics

Descriptive Statistics

Pig Meat

Quantile Statistics

Descriptive Statistics

Poultry Meat

Quantile Statistics

Descriptive Statistics

Milk Consumption

Quantile Statistics

Descriptive Statistics

Fish and Seafood

Quantile Statistics

Descriptive Statistics

Medical Expenditure

Quantile Statistics

Descriptive Statistics

Diphtheria

Quantile Statistics

Descriptive Statistics

Suicides

Quantile Statistics

Descriptive Statistics

NCD

Quantile Statistics

Descriptive Statistics

Distinct Count	1381
Unique (%)	66.6%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	11.6437
Minimum	0.26
Maximum	59.09
Zeros	0
Zeros (%)	0.0%
Negatives	0
Negatives (%)	0.0%

Distinct Count	726
Unique (%)	35.0%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	3.3526
Minimum	0
Maximum	55.4
Zeros	20
Zeros (%)	1.0%
Negatives	0
Negatives (%)	0.0%

Minimum	0
5-th Percentile	0.04
Q1	0.6
Median	1.445
Q3	3.49
95-th Percentile	13.027
Maximum	55.4
Range	55.4
IQR	2.89

Distinct Count	501
Unique (%)	24.2%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	1.584
Minimum	0
Maximum	22.31
Zeros	199
Zeros (%)	9.6%
Negatives	0
Negatives (%)	0.0%

Distinct Count	1307
Unique (%)	63.0%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	12.8389
Minimum	0
Maximum	64.24
Zeros	72
Zeros (%)	3.5%
Negatives	0
Negatives (%)	0.0%

Distinct Count	1522
Unique (%)	73.4%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	15.5803
Minimum	0.05
Maximum	72.74
Zeros	0
Zeros (%)	0.0%
Negatives	0
Negatives (%)	0.0%

Distinct Count	1980
Unique (%)	95.5%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	116.448
Minimum	1.02
Maximum	463.91
Zeros	0
Zeros (%)	0.0%
Negatives	0
Negatives (%)	0.0%

Distinct Count	1577
Unique (%)	76.0%
Missing	0
Missing (%)	0.0%
Infinite	0
Infinite (%)	0.0%
Memory Size	32.4 KB
Mean	17.395
Minimum	0.07
Maximum	191.75
Zeros	0
Zeros (%)	0.0%
Negatives	0
Negatives (%)	0.0%

Minimum	1.7014
5-th Percentile	2.8371
Q1	4.4335
Median	6.1969
Q3	8.1496
95-th Percentile	10.3428
Maximum	20.4134
Range	18.712
IQR	3.7161

Mean	6.3738
Standard Deviation	2.3791
Variance	5.6603
Sum	13219.2983
Skewness	0.3744
Kurtosis	0.2049
Coefficient of Variation	0.3733